Had companies whose large one-day fall came with a filing already trailed same-day fallers without one over the prior year?
Study 66 · Changed a rule or plan · recorded 23 October 2026 · answer: Not at the size tested · full paper with figures
The question
An earlier study found that falls with a company filing within a day kept falling from the next opening price while falls without one held. This second test of a programme on large one-day moves asks whether those companies had already trailed the market more over the prior year, as companies in trouble would.
Rules fixed before the result
Fixed in advance: the plain average decided unless its range exceeded 14 points; then the clipped average decided only if its range held the plain average and its false-alarm rate on the companies' ordinary years, with labels at random and with each company's own ups and downs, lay within 2.5% to 7.5% around the 5% target. The study would stop unread if overlapping years made the smallest detectable gap exceed ten points. Each year ends two sessions before the fall.
Data and method
Each company's share return against the market over the year before its fall, on the log scale, was compared between falls with a company filing within a day and falls without one on the same day and in the same price band; the filing's content was not read. The ranges allow for overlapping years and the same company recurring. The plain average decided; the clipped average's false-alarm rate was 5.8% (5.1% to 6.6%), within its allowed band.
The answer
Within the same day and price band, the prior year's return against the market of falls with a filing, less that of falls without one, was -1.4 points by the plain average (-5.5 to 3.0) and -1.4 by the clipped average (-5.5 to 2.8), so a ten-point shortfall is ruled out, though a smaller one remains possible. In the year after, a late filing notice followed falls with a filing 0.6 points less often (-2.7 to 1.5), inside the five-point size set in advance. The share of companies whose shares stopped trading within the year after differed by 0.03 points (-1.1 to 1.2). Falls without a filing were more volatile shares, and the log scale lowers a volatile share's yearly return. On summed daily returns, free of that lowering, falls with a filing trailed by 12.9 points (-20.8 to -5.1).
Main estimate
-0.014 fraction (95% range -0.055 to 0.030), from 15270 observations
What remains uncertain
These falls were studied before, so this is a check within our development data, not a test on unseen data. Condition here is the market's own pricing and the filing record, not the balance sheet; the window's first six months and newly listed companies are left out, and a gap under ten points is not ruled out. The log scale lowers a volatile share's year for the same simple return, and summed returns read the gap differently.
How it was made
Run took 90 minutes; independent code review: approve; the recorded calculation reruns to the same result.
Fingerprints
plan 8fbe068939c9789a836ec0acbac42062d19d61701c9bbbaf29e84b8b33f54ca5 result a9d7c0d32dc88f6404a49177e94269982d858f6e3b480019a7fd6d9c038772bb
Exploratory work on development data; not a performance claim.